Artificial Neural System Research Articles

Challenge of the performance management of trust control systems with deep learning

Introduction. The significance of machine learning under the conditions of digital transformation of industry, and methods of implementing deep learning to provide the performance of trust management systems are considered. The necessity of using convolutional artificial neural networks for deep machine learning is determined. Various technologies and architectures for the implementation of artificial neural networks are briefly considered; a comparative analysis of their performance is carried out. The work objective is to study the need to develop new approaches to the architecture of computing machines for solving problems of deep machine learning in the trust management system implementation.Materials and Methods. In the context of digital transformation, the use of artificial intelligence reaches a new level. The technical implementation of artificial neural systems with deep machine learning is based on the use of one of three basic technologies: high performance computing (HPC) with parallel data processing, neuromorphic computing (NC), and quantum computing (QC).Results. Implementation models for deep machine learning, basic technologies and architecture of computing machines, as well as requirements for trust assurance in control systems using deep machine learning are analyzed. The problem of shortage of computation power for solving such problems is identified. None of the currently existing technologies can solve the full range of learning and impedance problems. The current level of technology does not provide information security and reliability of neural networks. The practical implementation of trust management systems with deep machine learning based on existing technologies for a significant part of the tasks does not provide a sufficient level of performance.Discussion and Conclusions. The study made it possible to identify the challenge of the computation power shortage for solving problems of deep machine learning. Through the analysis of the requirements for trust management systems, the external challenges of their implementation on the basis of existing technologies, and the need to develop new approaches to the computer architecture are determined.

Using Artificial Neuro-Molecular System in Robotic Arm Motion Control-Taking Simulation of Rehabilitation as an Example.

Under the delicate control of the brain, people can perform graceful movements through the coordination of muscles, bones, ligaments, and joints. If artificial intelligence can be used to establish a control system that simulates the movements of human arms, it is believed that the application scope of robotic arms in assisting people’s daily life can be greatly increased. The purpose of this study is to build a general system that can use intelligent techniques to assist in the construction of a personalized rehabilitation system. More importantly, this research hopes to establish an intelligent system that can be adjusted according to the needs of the problem domain, that is, the system can move toward the direction of problem-solving through autonomous learning. The artificial neural molecular system (ANM system), developed early in our laboratory, which captured the close structure/function relationship of biological systems, was used. The system was operated on the V-REP (Virtual Robot Experimentation Platform). The results show that the ANM system can use self-learning methods to adjust the start-up time, rotation angle, and the sequence of the motor operation of different motors in order to complete the designated task assignment.

Characteristics of different swimming styles of swimming events based on artificial neural network data acquisition system

Characteristics of different swimming styles of swimming events based on artificial neural network data acquisition system

MODWT-ANN hybrid models for daily precipitation estimates with time-delayed entries in Amazon region.

Hydrological analyses based on precipitation records in the Amazon are essential due to their importance in climate regulation and regional and global atmospheric circulation. However, there are limitations related to data series with short periods and many gaps and failures at the daily scale. Thus, a hybrid model was developed based on an artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) coupled with the maximum overlap discrete wavelet (MODWT) method to obtain precipitation estimates. Six rainfall gauge stations located in different biomes within the studied region were adopted, and satellite data (CMORPH) were used. The interval of data that was have used is 1998-2016. The precipitation data were evaluated by seasonal (wet and dry) periods. The results obtained demonstrated the good capacity of the MODWT-ANFIS model to simulate the daily precipitation. In this case, data entries lagged by 4days and 5days performed better, with Nash values close to 1.0 and mean square errors (MSE) below 0.1.

Prediction of Peak Ground Acceleration by Artificial Neural Network and Adaptive Neuro-fuzzy Inference System

An attenuation relationship model belonging to a region with a high earthquake hazard is important. It is used for engineering studies to know how the peak ground acceleration (PGA) value depends on the distance where there are no stations. This study used earthquakes with magnitudes greater than 4 that IzmirNET recorded between 2009 and 2017 to determine the PGA through an artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS), which are widely applied in engineering seismology studies. For this purpose, 2925 records from 62 earthquakes were analysed in the ANN and ANFIS applications. Magnitude, focal depth, hypocentral distance (Rhyp), and site conditions comprise the inputs, and PGA values are the outputs. Using the Karaburun earthquake, we compared the ANN and ANFIS models using different ground motion prediction equations (GMPE) and the appropriate criteria. We determined the proximate values to PGA values measured at IzmirNET stations of the Karaburun earthquake, which was M = 6.2 in 2017, were used to test the ANN and ANFIS. The results were examined and indicated that the ANN and ANFIS are good candidates for obtaining PGA values for future earthquakes in the studied area. In addition, the PGA values of subsequent earthquakes can be calculated more quickly without any preliminary evaluation using an ANN and ANFIS.

Fuzzy Logic, Artificial Neural Network, and Adaptive Neuro-Fuzzy Inference Methodology for Soft Computation and Modeling of Ion Sensing Data of a Terpyridyl-Imidazole Based Bifunctional Receptor.

Anion and cation sensing aspects of a terpyridyl-imidazole based receptor have been utilized in this work for the fabrication of multiply configurable Boolean and fuzzy logic systems. The terpyridine moiety of the receptor is used for cation sensing through coordination, whereas the imidazole motif is utilized for anion sensing via hydrogen bonding interaction and/or anion-induced deprotonation, and the recognition event was monitored through absorption and emission spectroscopy. The receptor functions as a selective sensor for F- and Fe2+ among the studied anions and cations, respectively. Interestingly, the complexation of the receptor by Fe2+ and its decomplexation by F- and deprotonation of the receptor by F- and restoration to its initial form by acid are reversible and can be recycled. The receptor can mimic various logic operations such as combinatorial logic gate and keypad lock using its spectral responses through the sequential use of ionic inputs. Conducting very detailed sensing studies by varying the concentration of the analytes within a wide domain is often very time-consuming, laborious, and expensive. To decrease the time and expenses of the investigations, soft computing approaches such as artificial neural networks (ANNs), fuzzy logic, or adaptive neuro-fuzzy inference system (ANFIS) can be recommended to predict the experimental spectral data. Soft computing approaches to artificial intelligence (AI) include neural networks, fuzzy systems, evolutionary computation, and other tools based on statistical and mathematical optimizations. This study compares fuzzy, ANN, and ANFIS outputs to model the protonation-deprotonation and complexation-decomplexation behaviors of the receptor. Triangular membership functions (trimf) are used to model the ANFIS methodology. A good correlation is observed between experimental and model output data. The testing root mean square error (RMSE) for the ANFIS model is 0.0023 for protonation-deprotonation and 0.0036 for complexation-decomplexation data.

Machine Learning for the Estimation of Diameter Increment in Mixed and Uneven-Aged Forests

Estimating the diameter increment of forests is one of the most important relationships in forest management and planning. The aim of this study was to provide insight into the application of two machine learning methods, i.e., the multilayer perceptron artificial neural network (MLP) and adaptive neuro-fuzzy inference system (ANFIS), for developing diameter increment models for the Hyrcanian forests. For this purpose, the diameters at breast height (DBH) of seven tree species were recorded during two inventory periods. The trees were divided into four broad species groups, including beech (Fagus orientalis), chestnut-leaved oak (Quercus castaneifolia), hornbeam (Carpinus betulus), and other species. For each group, a separate model was developed. The k-fold strategy was used to evaluate these models. The Pearson correlation coefficient (r), coefficient of determination (R2), root mean square error (RMSE), Akaike information criterion (AIC), and Bayesian information criterion (BIC) were utilized to evaluate the models. RMSE and R2 of the MLP and ANFIS models were estimated for the four groups of beech ((1.61 and 0.23) and (1.57 and 0.26)), hornbeam ((1.42 and 0.13) and (1.49 and 0.10)), chestnut-leaved oak ((1.55 and 0.28) and (1.47 and 0.39)), and other species ((1.44 and 0.32) and (1.5 and 0.24)), respectively. Despite the low coefficient of determination, the correlation test in both techniques was significant at a 0.01 level for all four groups. In this study, we also determined optimal network parameters such as number of nodes of one or multiple hidden layers and the type of membership functions for modeling the diameter increment in the Hyrcanian forests. Comparison of the results of the two techniques showed that for the groups of beech and chestnut-leaved oak, the ANFIS technique performed better and that the modeling techniques have a deep relationship with the nature of the tree species.

Outcomes of Industry–University Collaboration in Open Innovation: An Exploratory Investigation of Their Antecedents’ Impact Based on a PLS-SEM and Soft Computing Approach

The outcomes of industry–university collaboration, in an open innovation context, may be of great support to firms, in their response to the challenges of today’s highly competitive environment. However, there is no empirical evidence on how these outcomes are influenced by their antecedents. Aiming to fill this gap, a research model to investigate the impact of the major antecedents, identified in the literature as motives, barriers and knowledge transfer channels on the beneficial outcomes and drawbacks of open innovation between the two organizations was developed in this study. The research model was empirically assessed, using a dual-stage predictive approach, based on PLS-SEM and soft computing constituents (artificial neural networks and adaptive neuro-fuzzy inference systems). PLS-SEM was successfully used to test the hypotheses of the research model, while the soft computing approach was employed to predict the complex dependencies between the outcomes and their antecedents. Insights into the relative importance of the antecedents, in shaping the open innovation outcomes, were provided through the importance–performance map analysis. The findings revealed the antecedents that had a significant positive impact on both the beneficial outcomes and drawbacks of industry–university collaboration, in open innovation. The results also highlighted the predictor importance in the research model, as well as the relative importance of the antecedent constructs, based on their effects on the two analyzed outcomes.

Comparative Study of DGA Based Fault Diagnosis using ANN and Fuzzy Systems

Dissolved Gas Analysis (DGA) method for fault detection has been implemented using Artificial Neural Networks (ANN), Fuzzy Logic (FL) and Adaptive Neuro Fuzzy Inference System (ANFIS). Incipient faults can be detected using DGA which provides reasonably good results. We have tried to improve this method in order to surpass its limitations. Comparative analysis using the mentioned methods have been done on IEC 599 standard, Rogers Ratio Method and Doernenburg’s method. Using Fault databases, the training has been done to improve the diagnostic capabilities. The obtained results clearly show the superiority of ANFIS on ANN and FL. Being a combination of both, its degree of accuracy in prediction and ease of use, provides a promising alternative in replacing the conventional methods.

Prediction Model for Geologically Complicated Fault Structure Based on Artificial Neural Network and Fuzzy Logic

The development and distribution of geologically complicated fault structure have the characteristics of uncertainty, randomness, ambiguity, and variability. Therefore, the prediction of complicated fault structures is a typical nonlinear problem. Neither fuzzy logic method nor artificial neural network alone can solve this problem well because the fuzzy method is generally not easy to realize adaptive learning function, and the neural network method is not suitable for describing sedimentary microfacies or geophysical facies. Therefore, taking the marginal subsags in the Jiyang Depression, Eastern China, as a study case, this paper uses the method of combining artificial neural network and fuzzy logic to study geologically complicated fault structure prediction model. This paper expounds on the research status and significance of geologically complicated fault structure prediction model, elaborates the development background, current status, and future challenges of artificial neural networks and fuzzy logic, introduces the method and principle of fuzzy neural network structure and fuzzy logic analysis algorithm, conducts prediction model design and implementation based on fuzzy neural network, proposes the learning algorithm of fuzzy neural network, analyzes the programming realization of fuzzy neural network, constructs complicated fault structure prediction model based on the artificial neural network and fuzzy logic, performs the fuzzy logic system selection of complicated fault structure prediction model, carries out the artificial neural network structure design of complicated fault structure prediction model, compares the prediction effects of the geologically complicated fault structure model based on artificial neural networks and fuzzy logic, and finally discusses the system design and optimization of the prediction model for geologically complicated fault structures. The study results show that the fuzzy neural network fully integrates the advantages of artificial neural network and fuzzy logic system; based on the clear physical background of fuzzy logic system, it effectively integrates powerful knowledge expression ability and fuzzy reasoning ability into the network knowledge structure of neural network, which greatly improves the prediction accuracy of geologically complicated fault structure.

The Performance Prediction of Electrical Discharge Machining of AISI D6 Tool Steel Using ANN and ANFIS Techniques: A Comparative Study

AISI-D6 steel is widely used in the creation of dies and molds. In the present paper, first the electrical discharge machining (EDM) of the aforementioned material is performed with a testing plan of 32 trials. Then, artificial neural networks (ANN) and adaptive neuro-fuzzy inference system (ANFIS) were applied to predict the outputs. The effects of some significant operational parameters—specifically pulse on-time (Ton), pulse current (I), and voltage (V)—on the performance measures of EDM processes such as the material removal rate (MRR), tool wear ratio (TWR), and average surface roughness (Ra) are extracted. To lead the process operators, process plans (i.e., parameter–effect correlations) are created. The outcomes exposed the upper values of pulse on-time caused by higher amounts of MRR and Ra, and likewise lower volumes of TWR. Furthermore, growing the pulse current resulted in upper volumes of the material removal rate, tool wear ratio, and surface roughness. Besides, the higher input voltage resulted in a lower amount of MRR, TWR, and Ra. The estimation models developed by using experimental data recounting MRR, TWR, and Ra. The root means the square error was used to determine the error of training models. Furthermore, the estimated outcomes based on the models have been proven with an unseen validation set of experiments. They are found to be in decent agreement with the experimental issues. The investigation shows the powerful learning capability of an ANFIS model and its advantage in terms of modeling complex linear machining processes.

AN APPLICATION OF SOFT COMPUTING TECHNIQUES TO PREDICT DYNAMIC BEHAVIOUR OF MOORING SYSTEMS

A spread mooring system (SMS) allows a ship or a floating platform to moor the seafloor using multiple mooring lines at a restricted region with a fixed heading in harsh weather. These systems can be used for the operations of ships of different tonnage at different sea depths. The optimal design of these systems is a challenging engineering problem because of the effects of many design parameters and changing environmental conditions. Modern soft computing techniques allow difficult engineering problems to be solved easily and precisely and are becoming more and more popular. In this paper, Artificial Neural Network (ANN) and Adaptive Neuro-Fuzzy Inference System (ANFIS) as soft computation techniques have been chosen to estimate the hawser tensions and displacements of a spread mooring system. The attained results show both techniques can give consistent indicators for the modelling of dynamic systems. Although these techniques performed very well, the ANFIS model is relatively superior to the ANN technique, considering the accuracy of hawser tensions and displacements in terms of the relative errors and coefficient of correlation obtained for the ANN and ANFIS.

A Comparative Analysis for Optical Character Recognition for Text Extraction from Images Using Artificial Neural Network Fuzzy Inference System

Artificial neural networks (ANN) has the capability to analyze raw data from processing input-output relationships. This function considers them important in areas of industry with such information is unusual. Researchers have tried to extract the information embedded within ANNs as set of rules used with inference systems to resolve the black-box function of ANNs. When ANN applied within a fuzzy inference system, the extracted rules yield high classification accuracy. In this paper a Multi-Layer Neural Feed-Forward Network using Artificial Neural Network Fuzzy Inference System (MLNFFN-ANNFIS) is proposed for accurate character recognition from images. The technique targets areas of business that have less complicated issues about which there is no simpler approach is desired to a complex one. This paper proposed an Optical Character Recognition model for Text Extraction from Images using Artificial Neural Network Fuzzy Inference System for accurate text detection from images. The technique proposed is more effective and simple than most of the techniques previously proposed. The proposed model is compared with various traditional models and the results indicate that the proposed model accuracy is more and performance is also improved.

Sistem Penyiraman Otomatis Tanaman Semusim Berbasis Jaringan Saraf Tiruan Multilayer Perceptron

In general, farmers water plants when the conditions are met, such as dry soil, no rain, and cold temperatures. One of the efficient ways to control it is to use an artificial neural network-based automatic plant watering system. The purpose of this study was to determine the success of artificial neural networks as decision-makers to water plants automatically. The stages of designing an automatic watering system based on an artificial neural network were to build software including artificial neural network modeling and Arduino microcontroller programming, automatically watering tools, evaluating tool performance, and testing tools in real-time. The test results show that the artificial neural network-based automatic plant watering system can water plants according to the given input pattern. The artificial neural network structure obtained is three neurons in the input layer, eight neurons in the hidden layer, and one neuron in the output layer. The artificial neural network-based automatic plant watering system succeeded in automatically watering two areas of land that the success rate is a 100%.Keyword— Automatic Watering, Microcontroller, ANN, Annual Crops.

Data-Driven Machine Learning Intelligent Tools for Predicting Chromium Removal in an Adsorption System

This study investigates chromium removal onto modified maghemite nanoparticles in batch experiments based on a central composite design. The effect of modified maghemite nanoparticles on the adsorptive removal of chromium was quantitatively elucidated by fitting the experimental data using artificial neural network (ANN) and adaptive neuro-fuzzy interference system (ANFIS) modeling approaches. The ANN and ANFIS models, relating the inputs, i.e., pH, adsorbent dose, and initial chromium concentration to the output, i.e., chromium removal efficiency (RE), were developed by comparing the predicted value with that of the experimental values. The RE of chromium ranged from 49.58% to 92.72% under the influence of varying pH (i.e., 2.6–9.4) and adsorbent dose, i.e., 0.8 g/L to 9.2 g/L. The developed ANN model fits the experimental data exceptionally well with correlation coefficients of 1.000 and 0.997 for training and testing, respectively. In addition, the Pearson’s Chi-square measure (χ2) of 0.0004 and 0.0673 for the ANN and ANFIS models, respectively, indicated the superiority of ANN over ANFIS. However, a small discrepancy in the predictability of the ANFIS model was observed owing to the fuzzy rule-based complexity and overtraining of data. Thus, the developed models can be used for the online prediction of RE onto synthesized maghemite nanoparticles with different sets of input parameters and it can also predict the operational errors in the system.

Bulk density prediction while drilling vertical complex lithology using artificial intelligence

Because of porosity and mineralogy disparity, rocks have a broad range of bulk densities, a critical property that has an important role in rock characterization. Practically, the density of rock is estimated either via wireline logging or logging while drilling tools. However, these techniques are not always accessible, making the door open for using different predicting methods such as empirical correlations. The existing empirical correlations have substantial restrictions, which limited their accuracy and reliability. This work targets developing various artificial intelligence models to predict bulk density for complex lithology during drilling (i.e., at real-time). The artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) techniques are used with the mechanical drilling parameters as inputs. This study uses a vertical well with 2912 data points from various lithologies including sand, carbonate, and shale to develop the models. Besides, a new empirical correlation for bulk density is developed based on the optimized ANN technique. A different dataset from the same tested field was used to validate the developed models. The obtained outcomes demonstrated that both ANN- and ANFIS-based models estimated the rock density with high fitting accuracy. The ANFIS approach results correlation coefficient (R) values of 0.99 with an average absolute percentage error (AAPE) values of 0.83 and 0.92% in training and testing processes. Whereas the ANN approach is slightly outperformed, as described by the highest R values of 0.99 and the lowest AAPE of 0.77 and 0.93% for training and testing processes. Also, a new empirical model for bulk density estimation was derived from the developed, optimized ANN-based model, which is applicable to the extracted weights and biases. The validation process indicates the robustness and reliability of the developed models with R values of 0.99 and AAPE of 0.92 and 0.96% for ANN- and ANFIS-based models, respectively. The outcomes of this work develop models with high accuracy for real-time prediction of bulk density, which can help in rock description in an adequate method with less cost, time, and errors.

Textile wastewater heavy metal removal using Luffa cylindrica activated carbon: an ANN and ANFIS predictive model evaluation

This study investigated the application of soft computing models [Artificial neural network (ANN) and Adaptive neuro-fuzzy inference system (ANFIS)] in removing heavy metals [chromium (VI), vanadium (V) and iron (II)] from textile wastewater using Luffacylindrica activated carbon (LAC). The effect of pH, contact time and adsorbent dosage on the adsorptive potential of the prepared LAC were determined using a batch mode experiment. Fourier Transform Infrared Spectroscopy and scanning electron micrograph assessed the potential of the adsorbent in this study. ANN and ANFIS were evaluated using the coefficient of determination (R2) and mean square error (MSE). The result showed that the models demonstrated significant predictive behavior with R2 (9.9999E−1), MSE (5.985E−14) for chromium(VI) removal, R2 (9.9999E−1), MSE (2.33856E−13) for iron(II) removal and R2 (9.9999E−1), MSE (7.22197E−12) for vanadium(V) removal for ANN, while ANFIS predicted R2 (0.76305), MSE (0.037105) for chromium(VI) removal, R2 (0.67652), MSE (0.846) for iron(II) removal, R2 (0.22673), MSE (0.65925) for vanadium(V) removal. Sensitivity analysis carried out with ANFIS (exhaustive search) indicated that the parameters (time, pH and adsorbent dosage) significantly impact the heavy metal removal. Thus, this study shows that ANN and ANFIS are reliable tools for modelling heavy metal removal using LAC. The parameter results obtained are relevant in process design and control.

Estimating the size of unerupted teeth: Moyers vs deep learning

This study aimed to design a deep learning (DL) system for estimating the sum of the mesiodistal widths (MDWs) of unerupted mandibular canines and premolars in the mixed dentition period and to clarify its performance by comparing DL estimates with Moyers' table (MT) results. The training dataset was obtained from 974 patients with permanent dentition. On the 3-dimensional digital models, MDWs of the mandibular right teeth were measured using Ortho Analyzer software (3Shape, Copenhagen, Denmark). A system was designed that could predict the total width of the mandibular canines and premolars using the mandibular central, lateral incisor, and first molar MDWs. This artificial neural system had 5 layers (4 hidden and 1 output) and 886 neurons. The MDWs of the mandibular teeth were introduced to the DL system in the form of datasets. The DL system's predicted results for 100 randomly selected patients were compared with the probability values obtained from the MT. The estimation performance of the DL system for the unerupted mandibular canines and premolars was acceptable, with 49.5% accuracy. The success rate for the MT, in comparison, was 45.0%, with an error margin of 1.00mm. The DL system offers a potential alternative to current methods in estimating unerupted tooth size. The results of the DL system are expected to provide diagnostic support for mixed dentition analysis on dental casts.

Prediction of Almond Nut Yield and Its Greenhouse Gases Emission Using Different Methodologies

The evaluation of a production system to analyze greenhouse gases is one of the most interesting challenges for researchers. The aim of the present study is to model almond nut production based on inputs by employing artificial neural networks (ANNs) and adaptive neuro-fuzzy inference systems (ANFIS) procedures. To predict the almond nut yield with respect to the energy inputs, several ANN and ANFIS models were developed, evaluated, and compared. Among the several developed ANNs, a network with an architecture of 8-12-1 and a log-sigmoid, and a linear transfer function in the hidden and output layers, respectively, is found to be the best model. In general, both approaches had a good capability for predicting the nut yield. The comparison results revealed that the ANN procedure could predict the nut yield more precisely than the ANFIS models. Furthermore, greenhouse gas (GHG) emissions in almond orchards are determined where the total GHG emission is estimated to be about 2348.85 kg CO2eq ha−1. Among the inputs, electricity had the largest contribution to GHG emissions, with a share of 72.32%.

Can GPM IMERG Capture Extreme Precipitation in North China Plain?

Extreme precipitation events (EPE) often cause catastrophic floods accompanied by serious economic losses and casualties. The latest version (V06) of the Integrated Multi-satellite Retrievals for Global Precipitation Measurement (GPM IMERG) provides global satellite precipitation data from 2000 at a higher spatiotemporal resolution with improved quality. It is scientifically and practically important to assess the accuracy of the IMERG V06 in capturing extreme precipitation. This study evaluates the two widely used products of IMERG during 2000–2018, i.e., IMERG late run (IMERG-L) and IMERG final run (IMERG-F), in the densely populated and flood-prone North China Plain. The accuracy of the IMERG V06 is evaluated with ground measurements from rain gauge stations at multiple scales (hourly, daily, and seasonally). A novel target tracking method is introduced to extract three-dimensional (3D) extreme precipitation events, and the near-real-time uncalibrated PERSIANN-CCS (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks Cloud Classification System) and GSMAP (Global Satellite Mapping of Precipitation) satellite data are added to further evaluate IMERG’s performance during extreme precipitation. Finally, for flash flood events induced by extreme rainfall in the Hebei Province from 15 to 23 July 2016, the accuracy of capturing the event with IMERG-F and IMERG-L was verified. Results reveal that IMERG-F is better than IMERG-L at all investigated scales (hourly, daily, and seasonally), but the difference between the two products is less at higher time resolutions. Both products manifest decreased performance when capturing 3D extreme precipitation events, and comparatively, IMERG-F performs better than IMERG-L. IMERG-F exhibits a distinct discontinuity in extreme precipitation thresholds between land and ocean, which is a limitation of IMERG-F not documented in previous studies. Moreover, IMERG-L and IMERG-F are comparable at an hourly scale for some metrics, which is beyond the expectation that IMERG-F is notably better than IMERG-L. This study provides a scientific basis for the performance of satellite precipitation products and contributes to guiding users when applying global precipitation products.